ai-insecure-output-handling
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of markdown documentation. It provides a technical overview of OWASP LLM Top 10 vulnerabilities (specifically LLM05:2025
- Improper Output Handling) for educational, pentesting, and defensive purposes.
- [NO_CODE]: There are no scripts, executables, or configuration files provided. The analysis found no evidence of command execution, network exfiltration, or persistence mechanisms.
- [SAFE]: While the document mentions payload examples (such as XSS and markdown exfiltration), these are presented as static text for demonstration and do not target the agent's execution environment or instructions.
Audit Metadata